INTRODUCTION Functionally selective antagonists: The sequencing of the human genome has revealed that the GPCR superfamily represents one of the largest protein families in the human genome with 367 members (excluding the olfactory and tastant receptors;(Vassilatis et al., 2003). Family A[unreadable][unreadable]the rhodopsin-like GPCR's-represents the largest group with 274 estimated members (Kroeze et al., 2003b). GPCRs remain one of the most viable targets for drug discovery efforts because of their easy 'drugability1 (Hopkins and Groom, 2002;Armbruster and Roth, 2005). Indeed, when considering psychiatric drugs, nearly every psychiatric indication is populated by drugs which primarily target GPCRs. Thus, atypical antipsychotic drugs target 5-HT2A-serotonin and D2- dopamine receptors in addition to a multiplicity of other GPCRs (Roth et al., 2004b) leading to both sideeffects (Kroeze et al., 2003a) and enhanced therapeutic efficacies (Roth et al., 2004a;Davies et ai., 2005a). Although many antidepressants preferentially target neurotransmitter transporters, they also have substantial 'off-target'actions at 5-HT2A and 5-HT2C receptors (Palvimaki et al., 1996;Dziedzicka-Wasylewska et al., 2006). As well, the atypical antidepressants (e.g. mirtazepine, mianserin, nefazadone) exert their actions via interactions with 5-HT2-family receptors and possibly other receptors (see http://pdsp.med.unc.edu/kidb.php ). Finally, several antidepressant, anti-insomnia, anti-anxiety and antipsychotic drugs currently in development target mainly GPCRs[unreadable][unreadable]particularly 5-HT2A receptors where they function as inverse agonists (Roth et al., 2004b;Roth and Kroeze, in press;Roth, 2006). According to classical concepts of pharmacology when inverse agonists are chronically administered upregulation of receptor number and super-sensitivity occurs (Urban et al., 2007). 5-HT2-family serotonin receptors are unusual in that inverse agonists paradoxically induce internalization and down-regulation in vitro and in vivo (Peroutka and Snyder, 1980;Blackshear and Sanders-Bush, 1982;Berry et al., 1996;Willins et al., 1999;Bhatnagar et al., 2001). A few years ago, we reviewed this general area and found nearly 100 examples wherein 5-HT2A antagonists of varying specificities induced receptor down-regulation in vivo (Gray and Roth, 2001). The mechanism responsible for receptor down-regulation appears to be receptor internalization and subsequent degradation (Willins et al., 1999;Bhatnagar et al., 2001) and not due to changes in receptor synthesis (Newton and Elliott, 1997;Roth and Ciaranello, 1991). For many GPCRs receptor internalization occurs via one of two distinct pathways: (a) internalization via endosomes (Ferguson et al., 1996) or (b) via caveolae (Roettger et al., 1995). We have found internalized 5-HT2A receptors to be associated with both endosomes (Berry et al., 1996;Bhatnagar et al., 2001) and caveolae (Bhatnagar et al., 2004), although the extent to which each pathway is used likely depends on the cellular milieu. The classical pathway for GPCR down-regulation involves arrestin binding and then translocation of the arrestin-receptor complex to clathrin-coated pits followed by internalization and either recycling to the plasma membrane or down-regulation (Ferguson, 2001);(Gray and Roth, 2002);Fig 1). Many proteins are involved in the post-endocytic decision-making process including GASP, SNX-1 and ubiquitin ligases. For 5-HT2A receptors it is clear that internalization may occur via both arrestin-sensitive and insensitive pathways (Bhatnagar et al., 2001;Gray et al., 2001b;Hanley and Hensler, 2002;Gray et al., 2003) in vitro although which pathways are essential in vivo is unknown. A major goal of this application is to understand how the ability of various 5-HT2A antagonists to induce internalization and down-regulation in vitro and in vivo affects their ability to treat schizophrenia and related disorders. As most of the previous preclinical studies were performed with psychiatric medications which target other GPCRs and neurotransmitter transporters (Gray and Roth, 2001) it is impossible to ascribe any beneficial action to effects on 5-HT2A receptor expression. We will evaluate relatively selective 5-HT2 antagonists which differ in their abilities to induce receptor internalization and down-regulation. Extensive preclinical studies indicate that 5-HT2A antagonists may prove beneficial in conditions of dopamine excess (Barr et al., 2004;Pehek et al., 2001) and NMDA-receptor hypofunction (Varty et al., 1999) as has been postulated to occur in schizophrenia. Salutary effects on excessive dopamine release are likely mediated via both midbrain and cortical 5-HT2A receptors (Nocjar et al., 2002;Pehek et al., 2006) while beneficial effects on glutamatergic function are likely mediated via direct actions on pyramidal neurons (Willins et al., 1997, Martin-Ruiz et al., 2001, Gonzalez-Maeso et al., 2007). Clinically, mixed results of 5-HT2Apreferring antagonists in schizophrenia have appeared. Early promising development of ritanserin (a 5- HT2A/2C antagonist) (Strauss and Klieser, 1991;Krystal et al., 1992) was halted due to cardiovascular safety issues. A large study, which has never been completely published, with M100907 showed beneficial effects greater than placebo but less than haloperidol (de Paulis, 2001;Potkin et al., 2001). A more recent study with SR46349B showed effects greater than placebo and equivalent to haloperidol (Meltzer et al., 2004). These equivalent results are reminiscent of a large trial comparing several atypical antipsychotic drugs and perphenazine (Lieberman et al., 2005). Currently, several 5-HT2A antagonists are in clinical trials as standalone or add-on medications for schizophrenia of which ACP-103 is furthest along. Phase II results for ACP- 103 demonstrated that co-therapy with low-dose risperidone was (a) more efficacious that risperidone alone; (b) was faster in onset;and (c) was associated with fewer side-effects (http://www.acadiapharm. com/programs/schizophrenia.htm ) as predicted by preclinical studies (Bayes et al., 2006;Vanover et al., 2006) Clinical trial data with selective 5-HT2A antagonists (SR46349B, M100907, ACP-103) as standalone therapies also reveals no significant cardiovascular or metabolic side-effects (Meltzer et al., 2004;Potkin et al., 2001)[unreadable][unreadable]actions which distinguish them from all currently available atypical antipsychotic drugs. Intriguingly, SR46349B, along with MDL11,939, are among the only 5-HT2A antagonists which cause receptor supersensitivity and up-regulation in vitro and in vivo with chronic administration (Rinaldi-Carmona et al., 1993;Rinaldi-Carmona et al., 1994;Gray and Roth, 2001;Aloyo et al., 2001;Harvey et al., 2004). SR46349B is also the only known 5-HT2A antagonist which has been demonstrated to be effective in a large-scale double-blind placebo controlled trial, suggesting that its unique ability to induce receptor up-regulation might be responsible for its clinical utility. It is unknown to what extent other selective 5- HT2A antagonists induce receptor redistribution in vivo although our prior studies with M100907 indicated that it efficiently promotes receptor internalization. As previously mentioned, receptor regulation is profoundly modulated by cellular milieu (Gray and Roth, 2001;Gray et al., 2001 a) and it is important that we will be profiling our compounds in native environments in vitro and in vivo. In this project we will examine the hypothesis that 5-HT2A inverse agonists which differ in their abilities to induce receptor redistribution in vivo will also differ in antipsychotic efficacy in various in vivo models. Because 5-HT2A receptors modulate the activity of glutamatergic pyramidal neurons (Aghajanian and Marek, 1997) (Martin-Ruiz et al., 2001) and cortical dopaminergic inputs (Nocjar et al., 2002;Pehek et al., 2005) either up- or down-regulation of 5-HT2A receptor number could conceivably have salutary effects on psychosis (Roth and Xia, 2004) and cognition (Harvey, 2003;Harvey et al., 2004). Designer receptors to probe the in vivo relevance of functional selectivity: It is likely that the actions of a functionally-selective ligand ultimately depends on the relative ability of a ligand to activate signal transduction pathways in discrete neuronal populations. Thus, in Fig 2 (top panels) different partial agonists have the ability to differentially activate the same receptor to a different extent depending upon the cellular milieu. In the panel on the left, Partial agonist "A" selectively activates the blue neurons and the summed action would reflect to a somewhat greater extent the actions of these neurons (A10 dopamine neurons for example). In the panel on the right, Partial agonist "B" selectively activates the green neurons and the summed activity would reflect to a somewhat greater extent the actions of these neurons (A9 dopamine neurons, for example). As can be seen, however, it is likely that any given drug will have some activity at both neuronal populations and that the summed behavioral activity will represent an integration response between the two neuronal populations. Currently, there is no technology which allows us to distinguish to what extent the behavioral actions of a functionally-selective ligand depend upon the discrete activation of different neuronal populations. The approach we've taken has been to create a technological platform which will allow us to, ultimately, precisely regulate the activity of discrete neuronal populations via drug-like compounds. What we have begun to do is to design families of receptors which can only be activated by pharmacologically-inert ligands and then to selectively express them in discrete neuronal populations. In the case above, these engineered receptors would induce a response upon activation that was solely 'blue'or 'green'in color. One could imagine, for instance, selectively modulating the activity of A9 or A10 dopamine neurons and determine the functional readouts of these sorts of activities. For this project, we will provide proof-of-concept studies whereby chemical switchs which selectively modulates the activity of single neurons or populations of neurons in vivo will be created.